Downloads · 30 days
136
36% of all-time downloads
granite-4.2-8b-4bit-mlx
brijrvl-ig/granite-4.2-8b-4bit-mlx
granite-4.2-8b-4bit-mlx is a text generation model from brijrvl-ig. Use it when you need the model to write or continue text. It is set up for mlx. The card lists the license as apache-2.0.
4-bit quantized MLX format of ibm-granite/granite-4.2-8b optimized for Apple Silicon inference.
All-time downloads
375
Public
Parameters
8.8B
4.9 GB on disk
Likes
3
Public
Repo makeup
Click a slice to open those files.
.safetensors4.9 GB · 100%
Parameter types
How the weights are stored.
U328.8B · 100%
Try a prompt
Model card
From the Hugging Face model README
granite-4.2-8b-4bit-mlx
4-bit quantized MLX format of ibm-granite/granite-4.2-8b optimized for Apple Silicon inference.
Features
- 4-bit Quantization: Low memory footprint (~4.9 GB), fast prefill and generation on Apple Silicon.
- Thinking / Chain-of-Thought: Native reasoning support via
<think> ... </think>tags. - Tool Calling: Tool use and function calling capabilities.
- Languages: Multilingual support across 12+ languages.
Usage with mlx-lm
pip install mlx-lm
from mlx_lm import load, generate
model, tokenizer = load("brijrvl-ig/granite-4.2-8b-4bit-mlx")
prompt = "What is the capital of France?"
messages = [{"role": "user", "content": prompt}]
prompt = tokenizer.apply_chat_template(messages, add_generation_prompt=True)
response = generate(model, tokenizer, prompt=prompt, max_tokens=256)
print(response)
- Task
- Text Generation
- Library
- mlx
- Type
- granite
- License
- apache-2.0
- Languages
- en, de, es, fr, ja, pt
- Created
- Sep 14, 2026
- Updated
- Sep 14, 2026